Lee-Daly-Huang-Saad-Rodriguez-DeVries-Seifert-A-solution-in-search-of-problems-a-cognitive-tool-for-solution-mapping-to-promote-divergent-thinking

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=cjen20Journal of Engineering Design ISSN: Print Online Journal

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=cjen20

Journal of Engineering Design

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/cjen20

A solution in search of problems: a cognitive tool for solution mapping to promote divergent thinking

Jin Woo Lee, Shanna R Daly, Aileen Huang-Saad, Gabriella Rodriguez, Quinton DeVries & Colleen M Seifert

To cite this article: Jin Woo Lee, Shanna R Daly, Aileen Huang-Saad, Gabriella Rodriguez,Quinton DeVries & Colleen M Seifert (2021): A solution in search of problems: a cognitivetool for solution mapping to promote divergent thinking, Journal of Engineering Design, DOI:10.1080/09544828.2021.1887462

To link to this article: https://doi.org/10.1080/09544828.2021.1887462

Published online: 01 Mar 2021.

Submit your article to this journal

View related articles

View Crossmark data

a Department of Mechanical Engineering, California State University Fullerton, Fullerton, CA, USA;

b Department of Mechanical, University of Michigan, Ann Arbor, MI, USA;cDepartment of Biomedical Engineering, University of Michigan, Ann Arbor, MI, USA;dDepartment of Electrical Engineering, University of Michigan, Ann Arbor, MI, USA; e School of Information, University of Michigan, Ann Arbor, MI, USA;

f Department of Psychology, University of Michigan, Ann Arbor, MI, USA

ABSTRACT

Engineering design processes are often defined as beginning with

a problem and diverging to generate possible solutions; however,

design processes can start with a newly developed technological

solution, followed by a divergent search for potential problem

appli-cations it can solve, termed ‘solution mapping’ Building on previous

research where engineering practitioners described their successful

strategies for solution mapping, we created a tool to support

solu-tion mapping and tested its impact with engineering students In

a single session, graduate and advanced undergraduate

engineer-ing students were presented with a novel technology and worked

to identify potential problem applications for it Comparing students

using the Solution Mapping Design Tool to two control groups,

more diverse problem applications were produced when using the

tool Considering diverse options is an important feature of design

processes shown to promote creativity and innovation With this

successful proof of concept, future work on solution mapping will

identify how to support engineers seeking problem applications by

making use of new technologies.

ARTICLE HISTORY

Received 9 July 2019 Accepted 4 February 2021

KEYWORDS

Design education; Design education, types of design; Design theory and research methodology, creative education; Design education, descriptive models of the design process; Design theory and research methodology, problem solving techniques; Creativity and innovation

1 Introduction

Prescriptive engineering design processes documented in the literature define an initialproblem as a starting point (Cross2008; Dym and Little2009; Dieter and Schmidt2009),proceeding next to understand the context and the needs of the stakeholders, generatepotential solutions, select options for prototyping, conduct testing and refining, and evalu-ate the solution (Kilgore et al.2007; Studer et al.2018) However, a problem is not always aninitial focus in engineering design Engineers who develop a novel technology may ‘reverse’this problem-first design process by beginning with the solution and exploring potential

problem applications In this ‘flipped’ process, called solution mapping, the engineer aims

to match a technological solution with a problem application, exploring multiple potential

CONTACT Jin Woo Lee jinlee@fullerton.edu Department of Mechanical Engineering, California State University Fullerton, 800 N State College Blvd., E-100, Fullerton, CA, USA

© 2021 Informa UK Limited, trading as Taylor & Francis Group

problems that can be solved with a new technology (Lee2019; Lee et al.2018a; Lee et al.2020a; Lee et al.2020b) For newly-developed technologies, identifying problems they canaddress is not obvious Shane (Shane2000) noted that although many people are exposed

to information about a new invention, very few opportunities to use a technology are covered, and these are not easily identifiable from simple knowledge of the technology.These challenges suggest the need for design tools and training, but engineering designcurrently lacks both for solution mapping

dis-Successful design tools have been developed to support other design processes, such

as concept generation in problem-first processes, and studies have demonstrated the cacy of leveraging design tools to aid engineers (Daly et al.2016; Hernandez et al.2013; Lee

effi-et al.2018c; Lee et al.2018d) Empirically-based design tool development has been guided

by analyses of patents and design artefacts (Lee et al.2018c; Altshuller1997; Camburn et al

2015) as well as by investigations of patterns in approaches to open-ended design tasks(Daly et al.2012; Lauff et al 2018) Many design tools have been developed to supportproblem-first design processes, but no evidence-based design tools are currently available

to guide engineers in solution mapping

Previous work on solution mapping has identified several cognitive strategies shared byengineering practitioners who successfully identified problem applications for new tech-nologies (Lee et al.2020a; Lee2019; Lee et al.2019) The use of cognitive strategies in theform of specific, experience-based guidelines to support decision making have been shown

to be representative of expertise (Sternberg et al.2003) The Solution Mapping Design Toolfor divergent thinking about technology problem applications was developed (Lee et al.2020a; Lee et al.2019) Divergent thinking, a necessary component of creativity (Kudrowitzand Dippo2013; Guilford 1967; Baer 2014), is defined as considering alternatives, mak-ing unexpected combinations, and identifying connections among remote associations(Guilford1967) (Treffinger et al.2002) In design, practicing divergent thinking to generatemultiple, diverse ideas has been shown to increase the potential for innovative outcomes(Brophy2001) In this study, we investigated the impact of the Solution Mapping DesignTool on divergent thinking when searching for technology problem applications

2 Related work

2.1 Solution first processes

Within engineering design, few studies have examined solution-first design processes.However, some studies have suggested a design process similar to solution mapping Forexample, a study of bio-inspired design observed that designers sometimes begin with amechanism seen in biological systems and search for engineering problems where it can

be applied (Helms et al.2009) Other studies describe a process starting with an existingproduct in the marketplace and, through ‘reverse engineering,’ identifying its mechanismsand making improvements based on the perceived market defects or envisioned evolution(Otto and Wood2001) The literature suggests that by analysing a product to identify itscomponents, reverse engineering can help better understand the needs of the application(Keller et al.1999) While these alternative design processes also begin with a known solu-tion, they retain some problem definition through the existing product or biological system

in the form of what the solution is designed ‘for.’

Literature in entrepreneurship has examined solution-first processes, described as

‘technology-push,’ approaches, where research and development of new technologies candrive product development (Di Stefano et al.2012) For example, Maine and Garnsey (Maineand Garnsey2006) described that generic technologies, such as new materials, could bedeveloped into a new product by matching them to needs in the market The entrepreneur-ship literature has described solution-first approaches more broadly as ‘opportunity recog-nition’ processes to identify a potential market or need Opportunity recognition reportshave mainly focused on the activity of business entrepreneurs aimed towards marketand stakeholder needs and the commercialisation process However, some studies defineopportunities with regard to new technology by beginning with a technological inventionand recognising an opportunity for its use and an approach to its exploitation (Shane2000;Grégoire et al.2009)

We have defined solution mapping as a more specific form of opportunity recognitionfocused on a newly developed technology without an evident purpose; through divergentthinking processes, possible problems to solve using the technology are identified (Lee

et al.2020a; Lee2019; Lee et al.2019) For example, an engineer developed an organicphotodetector as he focused on creating materials for electronics He later identified anapplication to create a lighting system that changes colour and intensity in real-time in syncwith the sun’s position, providing daylight in spaces deprived of windows and skylights(Allen2015) When a new technology is created, it is an opportunity to solve problems;most centrally to solution mapping, the technology lacks a general purpose, but has agreat fit to a constrained problem For example, the creation of a ‘not very sticky’ adhe-sive at 3M presented a generic opportunity as a fixative and it was later used to solve theproblem of marking places on paper This fixative on Post-It Notes was a classic example

of solution mapping: a specific technology is matched to a previously unrelated problemthrough a divergent design process leading to discovery This solution mapping design pro-cess clearly differs from problem-first design processes because the problems are initiallyundefined

2.2 Development of design tools and strategies for divergent thinking

Design tools and strategies have been developed to support divergent thinking in theearly stages of design, and successfully promote the consideration of multiple, diverseoptions leading to more creative and innovative outcomes (Lee et al.2018c; Daly et al

2012; Kudrowitz and Dippo2013; Brophy2001; Shah et al.2000; Linsey2007) Idea ation tools aimed primarily at problem-first processes have been developed using severaldifferent methods One approach in creating design tools derived from studying success-ful design outcomes; for example, TRIZ was developed by studying patterns in over 40,000patents to create a set of 40 strategies supporting concept generation (Altshuller1997).The effectiveness of TRIZ was examined in one ideation study where students improved ingenerating ideas with higher variety and novelty compared to a control group (Hernandez

gener-et al.2013) Similarly, a set of ideation strategies in microfluidics was developed by ing patterns evident in patents; in a study with advanced engineering students, the use ofthe microfluidic design strategies resulted in more varied and creative ideas during con-cept generation (Lee et al.2018c) These design tools have been proven to be effective atimproving fluency and flexibility when used in problem-first design scenarios

extract-Other design tools have been developed through empirical studies of designers’ workprocesses Synectics, a problem-solving methodology, was derived from audio and videorecording meetings dealing with obstacles and arriving at creative solutions (Gordon1961).Synectics emphasises problem-solving using analogies to generate solutions, as doesdesign-by-analogy (Linsey2007; Tomko et al.2015) Design-by-analogy was developed toguide designers in linguistically representing the design problem that can support diver-gent thinking by creating novel analogies and analogous domains (Linsey2007; Linsey et al.

2008; Fu et al.2015; Linsey et al.2012) However, design-by-analogy has been developedand tested to promote generating novel solutions for a problem, not for using a solution togenerate possible problem applications

While brainstorming is common in group design, it is sometimes practiced with its nal guidelines to generate as many ideas as possible, minimise evaluation while generating,and build on previous ideas (Osborn1963; Osborn1942) However, brainstorming does

origi-not provide specific directions for creating novel ideas Design Heuristics were developed

by identifying designers’ generation strategies as they worked through open-ended lems in think-aloud sessions (Daly et al.2012) and Design Heuristics have been documented

to support students and professional designers in generating creative solutions to lems (Yilmaz et al.2016; Lee et al.2018b) As designers verbalised their thought processes,researchers captured specific strategies used to introduce variation in the potential solu-tions generated This suggests that engineers who have successfully identified problemapplications of a novel technology may offer alternative or more specific strategies forsolution mapping

prob-Researchers in entrepreneurship have identified a variety of strategies for recognisingopportunities through empirical studies of entrepreneurs Baron and Ensley (Lee et al.2018b) described that entrepreneurs ‘connect the dots’ between seemingly unrelatedevents and detect meaningful patterns to identify new product opportunities However,the ‘connect the dots’ strategy provides limited guidance on how to identify meaning-ful patterns Arentz and colleagues (Baron2006) stated that leveraging prior knowledgedeveloped from past experiences can help in identifying opportunities This strategy is chal-lenging to use because it does not specify which prior knowledge should be considered.These strategies in entrepreneurship can benefit from having prescriptive guidance on how

to find potential problem applications for technologies

An educational approach to identify potential problem applications of technologieswas established by the National Science Foundation’s I-Corps program (Arentz et al.2013;Nnakwe et al.2018) I-Corps participants follow a curriculum developed by Steve Blank toinvestigate different uses and commercialisation potential for their technology (Robinson

2012) The standard process entails customer discovery to identify potential partners, andmeetings with business investors to gain insights about developing a viable product Inthe curriculum, participants are required to complete over 100 interviews with potentialstakeholders to understand the needs that their technology can fill The interviews serve

as a good opportunity to confirm or deny their assumptions about possible uses of theirtechnologies However, in the I-Corps program, there are limited strategies provided tosupport designers in forming initial assumptions about potential problem applications fortheir technology to support solution mapping Thus, we aimed to develop and examine theeffectiveness of a design tool to support solution mapping

To identify potential strategies for identifying problem applications for a new ogy, empirical evidence from successful solution mapping design cases were collected in arecent study (Lee2019; Lee et al.2020a; Lee et al.2020b) Through interviews with engineerswho developed novel technologies and successfully matched them to problem applica-tions, a set of cognitive strategies for solution mapping was identified Cognitive strategiesare specific experience-based guidelines identified in practices that appear helpful to goodoutcomes (Blank and Dorf2012), and using these strategies has been shown to be highlyadvantageous in diverse settings (Riel1996; Brown and Goslar1986; Lawson1979) Cog-nitive strategies in approaching design tasks have been developed into explicit designapproaches that can be adopted by others (Lee et al.2018c; Daly et al.2012; Brown andGoslar1986).

technol-Beginning with these empirically-derived guidelines, we formulated a design tool to port engineers as they attempted solution mapping The purpose of the tool is to promotedivergent thinking by helping engineers consider multiple, diverse problem applicationsfor a technology The tool was designed to follow the process observed in successful engi-neering designers working with new technological solutions: (1) break down the technol-ogy into key characteristics; (2) identify enabling functions based on these characteristics;(3) search in multiple industry sectors; and (4) identify specific needs (Lee et al.2020a; Lee

sup-2019; Lee et al.2019) While evident in expert solution mapping, no instruction on thisalternative design process or instructional strategies yet exists

To implement the tool, we created a solution mapping process within a single designsession Given a technology, the engineers are guided through a sequence of steps wherethey analyse the technology’s key characteristics as enabling functions Then, they con-sider a list of industry sectors in order to survey a broad range of problem application areas.Finally, they propose multiple problems where the technology may be of use The SolutionMapping Design Tool resulted from a series of pilot tests designed to ensure it was easy

to understand and apply to novel technologies It was also considered through several ferent technologies to ensure the tool’s language was suitably independent of a specifictechnology

dif-In the study, we compared the impact of the use of the Solution Mapping Design Tool

to two control groups One group was provided only the list of industry sectors also used

in the tool This comparison would allow us to determine whether the information aboutsectors was sufficient to motivate problem discovery apart from the tool The true controlgroup completed the same solution mapping task without any assistance

applica-3.1 Materials

The Solution Mapping Design Tool was developed from a study of 19 engineering fessionals (Lee et al.2020a; Lee2019; Lee et al.2019) who developed novel technolo-gies and then, identified one or more problem applications in a wide variety of industrysectors, including energy, biotechnology, aerospace, manufacturing, and materials Todevelop the tool, a graduate student trained in qualitative research methods conductedsemi-structured interviews with engineering professionals that aimed to identify cognitivestrategies used to find problem applications for technologies The engineering profes-sionals had experience being successful at developing novel technologies and identifyingproblem applications, with success defined as developing commercial products or receiv-ing external funding for commercialisation The semi-structured interviews were designed

pro-to examine specific patterns identified across individuals performing solution mapping.The semi-structured interview data revealed cognitive strategies in solution mappingand a sub-set of the cognitive strategies was translated into a two-part design tool Thefirst part of the tool provides scaffolding for exploring the technology (as shown in Figure1).The tool worksheet guides students in (1) identifying key characteristics of a technology, (2)examining its enabling functions, (3) identifying areas for potential use, and (4) designingspecific applications of a technology The second part of the tool includes industry infor-mation in the form of a page of listing industry sectors from the North American IndustryClassification System (as shown in Appendix A1) This two-part design tool was pilot-testedwith 15 engineering students and we iterated on its design multiple times to ensure clarityand usability The pilot test participants received $25 for approximately 1.5 h of their time.The pilot tests prompted additional iterations of the solution mapping tool to ensure clarityfor the users Thus, the results from the pilot tests were not included in the data analysis asthe tool itself was modified after the initial testing

3.2 Data collection

Data collection proceeded through four major stages as represented in Figure2

Students were recruited by email to participate in a single experimental session ofapproximately 1.5 h through engineering department listservs at a large Midwestern uni-versity A total of 93 engineering students (54 male and 39 female; 68 advanced undergrad-uate and 25 graduate level; and 63 in mechanical engineering and 30 in non-mechanical(e.g biomechanical engineering)) participated in the study All participants received $25 ascompensation for their participation

Students were assigned at random to one of three group sessions: (1) a control group,where no additional information was provided (N= 32); (2) a second control group where

only a list of Industry Sectors was provided (Appendix A1; N= 28), and (3) a Solution ping Design Tool group that was given both the list of Industry Sectors and scaffolding to

Map-guide the solution mapping process (N= 33) These groups were roughly proportional tothe demographics of the group as a whole

The students were directed to work individually to ‘identify diverse problem tions for a new technology and to generate as many problem applications as possible’ in

applica-60 min A graduate student with prior teaching experience conducted all the study sessionsthrough a PowerPoint presentation for consistency An introduction to solution mapping

Figure 1.Solution Mapping Design Tool Part 1.

Figure 2.Data collection overview

was presented for the first 5 min, and students had the opportunity to ask questions Allgroups were then directed to identify potential problem applications for ‘shape memoryalloys.’ This novel mechanical engineering technology was selected through a pretest toensure its accessibility for undergraduate engineering students This prompt was presented

to all students in the study:

3.3 Shape memory alloy

A shape-memory alloy is an alloy that remembers its original shape When it is deformed,

it can return to its pre-deformed shape when heated The transformation temperature can

be adjusted to be between−100°C to 200°C through changing the alloy composition The

two main types of shape-memory alloys are copper-aluminum-nickel and nickel-titanium.These compositions can be manufactured to almost any shape and size The yield strength

of shape-memory alloys is lower than that of conventional steel, but some compositionshave a higher yield strength than plastic or aluminum

Identify potential applications of shape memory alloys

Please spend 1 h to complete this task

Blank sheets were given to students to document the problem applications they ated, and students were prompted to both sketch and describe their problem applications(as shown in Figure3) The design sheets were collected at the end for analysis

gener-3.4 Data analysis

Drawings from the design sheets were scanned and the associated written descriptionswere transcribed The size of the drawings was adjusted to be similar for all designs to clearlypresent them to the coders An example of a student’s sheet with a transcribed description

is shown in Figure4

The analysis of each participant’s collection of ideas focused on two criteria associatedwith divergent thinking in design (Daly et al.2016; Navarro-Prieto et al.1999) and creativ-ity research (Guilford1967; Kudrowitz and Wallace2013; Amabile1982): (1) fluency (total

Figure 3.Concept sheets for individual problem application ideas

Figure 4.Example of a problem application with a design sketch and transcribed description.

quantity) and (2) flexibility (diversity of applications) To measure the quantity of ideas

in a participant’s set, we counted the total number of problem applications generated,indicated by the number of individual sheets they completed

We measured the diversity of each student’s collection of ideas using two means: (1)

functions, and (2) industry sectors We examined variations in how the shape memory alloys were used (functions) and where they were applied (industry sectors) The list of industry

sectors was modified to serve as codes to identify variations in application areas (AppendixA2) In examining diversity, each problem application was labelled with only one functionand industry sector Researchers quantified the number of unique functions and industrysectors considered by each participant The code list of functions was generated inductivelythrough multiple iterations of three coders For example, one student considered using theshape memory alloy to design an automobile body (as shown in Figure5) If dented, it could

be heated to remove the dent This problem application function was categorised as ‘motorvehicle and parts’ industry and ‘self-repair’ function The list of different functions for shapememory alloys generated during the analysis is shown in Appendix A3

Two independent coders then categorised every application design for the industrysector and function The percent agreement was 86% for industry sectors and 84% for func-tions, greater than the 70% level typically accepted for inter-rater reliability (Wilson et al

1954) The researchers discussed all discrepancies until there was a consensus

Figure 5.Example of an application of shape memory alloy that was categorised as ‘motor vehicle andparts’ industry and ‘self-repair’ function

A one-way analysis of variance (ANOVA) compared the three instructional groups on theoutcome measures of the quantity of application designs and diversity measures of industrysectors and functions considered ANOVA is used to analyse the differences among groupmeans in a sample (Osborne2008) The assumption of homogeneity of variance was notviolated and Tukey’s honestly significant difference (HSD) post hoc test was used Addi-tionally, we analysed the effects of major (mechanical versus non-mechanical engineeringstudents), gender (male versus female), and grade level (undergraduate versus graduatestudent) using Wilk’s lambda tests in multivariate analysis of variance (MANOVA) on the out-come measures of the quantity of applications, industry sectors, and functions MANOVAtests multiple dependent variables to examine the effects of the measured dependentvariables (Girden 1992) We used the error rate of alpha= 0.05 for both ANOVA andMANOVA.

4 Results

As a check on our assumptions, we compared the results for students divided based ondemographic groups There were no significant differences in either the quantity or diver-sity measures when comparing mechanical and non-mechanical engineering students

(p = 0.109), male and female students (p = 0.787), and advanced undergraduate and uate students (p= 0.183), across all three groups Thus, we present the results in thefollowing subsections by tool condition (full Solution Mapping Design Tool, the controlgroup with a list of industry sectors only, and the control group)